Solid state drives (SSDs) may wear out over time as the NAND technologies that the drives are based on have a limited number or program/erase cycles. Over time, SSDs may reach a point where an erase cycle fails to reset a NAND flash block to a writable state. Such an SSD may be said to be failed. Current disk array controllers designed to handle traditional spindle-based disk drives may employ failure modes that affect both reading and writing, rendering the drives useless.
However, some SSD may employ a read-only failure mode that allows the drive to execute read commands successfully even though write commands result in a failure. As the typical response of current disk array controllers to a failed write command is to designate the drive as failed, such controllers can not effectively utilize the read-only failure mode of SSDs.